Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

Page created by Duane Vazquez
 
CONTINUE READING
A picture of the National Audit Office logo

Report
by the Comptroller
and Auditor General

Cabinet Office and Department of Health & Social Care

Investigation into how
government increased
the number of ventilators
available to the NHS in
response to COVID-19

HC 731   SESSION 2019–2021   30 SEPTEMBER 2020
4 What this investigation is about
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

                             What this investigation is about

                             1     Ventilators are medical devices that assist or replace a patient’s breathing.
                             Patients with COVID-19 who are admitted to hospital often have problems
                             breathing. On arrival in hospital a patient’s blood oxygen level is measured. If it
                             is low, then the patient may be given: standard oxygen therapy using a mask;
                             non-invasive ventilation where oxygen is delivered under pressure via a mask or
                             helmet; or invasive mechanical treatment using a mechanical ventilator, which
                             takes over a patient’s breathing. Treatment is a judgement for clinicians and
                             patients may undergo more than one treatment during a stay in hospital.

                             2     Our investigation covers how government increased the number of
                             ventilators available to the NHS from March 2020, in response to the COVID-19
                             pandemic through:

                             •     Department of Health & Social Care (DHSC) purchasing of ventilators on
                                   the global market, as part of a wider DHSC and NHS England and NHS
                                   Improvement (NHSE&I) oxygen and ventilation programme; and

                             •     Cabinet Office’s ‘ventilator challenge’ to encourage UK businesses to design
                                   and manufacture more mechanical ventilators.

                             3     We explain:

                             •     what ventilators are and how they are used in treating COVID-19 (Part One);

                             •     government’s objectives and performance in increasing the number of
                                   ventilators available to the NHS (Part Two);

                             •     how DHSC purchased ventilators (Part Three); and

                             •     Cabinet Office’s ventilator challenge (Part Four).

                             4    This report mainly covers the period March 2020 to September 2020.
                             We conducted our fieldwork in the period June 2020 to August 2020. We focus
                             on the Cabinet Office and DHSC and NHSE&I programmes but also explain other
                             public bodies’ roles, including the Medicines and Healthcare products Regulatory
                             Authority. We consulted ventilator designers and manufacturers, doctors and
                             other experts. Appendix One sets out our methodology in more detail.
What this investigation is about 5
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

             5     This investigation is part of a programme of work the National Audit Office
             is undertaking to support Parliament in its scrutiny of government’s response
             to COVID-19. It focuses on how government used public money to increase the
             number of ventilators available to the NHS, in response to COVID-19. We do
             not look at how ventilators are used by the NHS. As such, we do not express
             any medical or technical opinion on the role of ventilators in treating COVID-19,
             or on the safety, effectiveness, functionality or any other aspect of ventilators’
             performance. Furthermore, a ventilator is only part of the resources required to
             treat COVID-19 patients. Hospitals also need: skilled staff; power; bed space; an
             oxygen supply; drugs; and additional equipment such as monitors and feeding
             pumps. A wide variety of consumable products such as filters are also required.
             We do not cover these wider factors in this report.

             6     This investigation examines the departments’ approach to deciding which
             ventilators to purchase, but we did not audit in detail each of the individual
             contracts and transactions entered into and we do not express any legal opinion
             on the use of public procurement regulations or indemnities granted by the
             departments involved. We are currently investigating government procurement
             during the COVID-19 pandemic, including the use of emergency procurement
             regulations more generally. We will publish a report on this later this year.
6 Summary
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

                             Summary

                             Key findings

                             The delivery of ventilators against estimated need
                             1     In the early stages of the pandemic the NHS believed it could need far more
                             mechanical ventilators than were available. By the beginning of March 2020,
                             NHS England and NHS Improvement (NHSE&I) modelling, based on reasonable
                             worst-case planning assumptions assured by the Scientific Advisory Group for
                             Emergencies (SAGE), indicated that the NHS could need up to 90,000 beds with
                             ventilators to care for COVID-19 patients. Meanwhile, NHSE&I’s survey of NHS
                             trusts in England, conducted in late February and early March, indicated that
                             the NHS only had access to an absolute maximum of around 7,400 mechanical
                             ventilators, including some that would not normally be used to treat adult
                             patients in a hospital bed, such as ventilators from ambulances and paediatric
                             departments (paragraphs 2.4 and 2.7 and Figure 2).

                             2    Government decided from 13 March to pursue all available options to
                             acquire as many ventilators as possible, as quickly as possible. This followed
                             DHSC’s initial efforts from 3 March to secure as many ventilators as possible
                             through existing routes. The government’s strategy was to:

                             •     buy as many ventilators as possible from both UK and global suppliers
                                   as part of a wider ‘oxygen, ventilation, medical devices and clinical
                                   consumables’ programme (with contracts let by DHSC as part of this
                                   wider joint programme with NHSE&I); and

                             •     encourage UK manufacturers to scale up production of mechanical
                                   ventilators (the ventilator challenge, led by Cabinet Office).

                             The two departments ran their programmes separately but worked towards the
                             same overall targets and exchanged data on their progress in acquiring ventilators
                             daily. On 16 March, government announced a “call to arms” to industry bodies to
                             provide as many ventilators as they could (paragraphs 2.8, 3.2, 3.3, 3.7, 3.8 and 4.7).
Summary 7
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

             3    Government acquired 1,800 new ventilators before the mid-April peak
             of the pandemic. By 24 March, NHSE&I had revised down its estimate of the
             number of ventilated beds that could be needed in England on 13 April to
             17,500, based on the latest SAGE-assured reasonable worst-case planning
             assumptions. However, in the week commencing 13 April, only around 10,900
             mechanical ventilator machines were available to the NHS across the whole UK.
             This comprised around 9,100 existing units the NHS had by then found it already
             had access to across the UK, around 1,200 on loan from the private sector,
             around 400 newly purchased by DHSC and around 200 manufactured through
             the Cabinet Office ventilator challenge (paragraphs 2.9 and 2.10, and Figure 3).

             4    In the event, the new ventilators were not needed at the April peak
             because demand was considerably lower than the reasonable worst-case
             scenario. NHSE&I data indicate that in total, around the peak of COVID-19
             hospital admissions on 14 April, NHS providers in England had 6,818 ventilator
             beds operational, of which: 2,849 were occupied by COVID-19 patients; 1,031
             were occupied by other patients; and 2,938 were unoccupied.1 DHSC and
             NHSE&I are not aware of any point when a patient who needed a ventilator was
             unable to get one. They allocated new ventilators to NHS trusts in England to
             meet local demand based on trusts’ individual requests, and analysis of trusts’
             current ventilator usage and patterns of demand (paragraphs 2.10, 2.12 and 2.18).

             5    On 15 April government adopted formal targets to aim for 18,000
             mechanical ventilators by the end of April and 30,000 by the end of June.
             Government deliberately set targets that were substantially higher than demand
             at that point, in order to build a surplus that:

             •      covered potential regional variations in demand and hospital stock; and

             •      provided a sufficient safety margin to be confident that the UK had enough
                    ventilators, including in the event of a possible second peak.

             Government did not consider it necessary to set targets for non-invasive
             ventilators (paragraph 2.14 and Figure 2).

             6    Government missed its target for the end of April. By the end of April
             government had increased the total number of mechanical ventilators to around
             11,500, meaning it missed its target by 6,500. However, the NHS still had more
             than the actual demand for ventilators (paragraph 2.15 and Figure 3).

             1   The 6,818 operational beds total is lower than the 7,400 ventilators available set out in paragraph 1 as at any
                 given time some ventilators would be unavailable due to cleaning and servicing.
8 Summary
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

                             7     Government was closer to meeting its end of June target of 30,000
                             ventilators, surpassing this number in early August. By 30 June government
                             had acquired around 24,000 mechanical ventilators against its target of
                             30,000. This comprised around: 9,100 already available to the NHS; 2,600
                             units purchased by DHSC and 12,300 built through Cabinet Office’s ventilator
                             challenge. At this point devices borrowed from the private sector had been
                             returned. The departments met the 30,000 target around 3 August. In addition,
                             by 9 July, the NHS had 27,700 non-invasive ventilators and continuous positive
                             airway pressure (CPAP) machines, including up to 17,800 purchased by DHSC
                             since March (paragraph 2.16 and Figure 3).

                             8    Most of the new ventilators are being held in reserve. As at 16 September
                             only around 2,150 mechanical ventilator units acquired through the programmes
                             had been dispatched to the NHS. This is because the anticipated demand
                             did not materialise. The remaining units, which were largely purchased
                             in case of increased demand in future waves of COVID-19, are stored in
                             warehouses, including the Ministry of Defence’s facility at Donnington, as a
                             central reserve. DHSC and NHSE&I told us that they are distributing some of
                             these devices to NHS trusts to prepare for potential future waves of COVID-19
                             (paragraphs 2.17 and 2.20).

                             The purchase of ventilators on the world market
                             9     DHSC had purchased all the stock of mechanical ventilators it could
                             from established NHS suppliers by 13 March. In early March, DHSC contacted
                             established ventilator suppliers operating in the UK and placed orders for
                             as many ventilators and other oxygen therapy devices as suppliers could
                             provide at the time. This increased the number of non-invasive ventilators and
                             oxygen concentrators available to the NHS but was less successful in securing
                             mechanical ventilators. DHSC explained that this was largely because mechanical
                             ventilators are more complex devices which take longer to build so suppliers do
                             not have large amounts of stock (paragraphs 3.3 to 3.4 and Figure 4).

                             10 From 17 March DHSC placed significant orders directly with overseas
                             manufacturers and their agents. Following government’s “call to arms” on
                             16 March, Cabinet Office and DHSC both received a large number of offers
                             from intermediary bodies who said they had access to ventilators built overseas.
                             DHSC found that these offers took a long time to review and that many offers
                             were not backed up with firm commitments of stock. Only one such offer led
                             to a purchase of ventilators (paragraphs 3.7 and 3.8, and Figure 4).
Summary 9
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

             11    As the programme progressed, DHSC decided to only deal directly
             with overseas manufacturers and their accredited distributors. It focused
             particularly on China as a large market with less coverage from the UK
             distributors on its existing frameworks. It worked with the China offices of the
             Foreign & Commonwealth Office and the Department for International Trade,
             which provided personnel and expertise to assess potential purchases and
             place orders as quickly as possible. It placed the vast majority of its orders
             for mechanical ventilators before government set its longer-term targets on
             15 April (paragraphs 3.7 to 3.10 and Figures 4 and 9).

             12 DHSC experienced increasing global competition to buy ventilators and
             made purchases primarily on the credibility of the offer, not price. It did not
             set a maximum price it was willing to pay but weighed up a number of factors
             in each case including: speed of delivery; the credibility of the supplier; and
             the clinical suitability of the devices. It found global competition for buying the
             ventilators intensified over time and that prices rose steadily as stock became
             harder to obtain. Suppliers usually insisted on payment upfront to secure prices
             and manufacturing slots. This meant DHSC accepted the risks both on the
             quality of devices, and that clinicians may not deem it appropriate to use devices
             for purposes other than their intended use. DHSC told us it only knows of one
             such issue, where 750 transport ventilators were bought, at a cost of around
             £2.2 million, that clinicians were not comfortable to use in an intensive care
             unit environment (paragraphs 3.10, 3.11 and 3.13).

             Cabinet Office’s ventilator challenge
             13 Cabinet Office needed to find mechanical ventilator designs that both
             worked and could be produced at scale. From 13 March it worked with industry to:

             •      develop new, or modify existing, ventilator or anaesthesia machine designs
                    to meet standards that the Medicines and Healthcare products Regulatory
                    Agency (MHRA) developed for rapidly manufactured ventilators; and

             •      increase manufacturing capacity to build each design at a much greater
                    scale than usual. This meant securing new factory capacity for each
                    design, managing global supply chains and ensuring regulatory approvals
                    were in place. Each manufacturer taking legal responsibility for the
                    ventilator needed to be subject to MHRA scrutiny, to consider whether an
                    “exceptional use” authorisation could be considered for devices that did
                    not already have ‘CE’ marking.

             MHRA told us that regulatory approval in normal circumstances could take
             18 to 24 months, so achieving approval for new or even modified designs,
             their production facilities and supply chains in a few weeks was a significant
             challenge (paragraphs 4.1 to 4.6).
10 Summary
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

                             14 MHRA updated the ventilator specification several times, reflecting
                             growing clinical experience of treating COVID-19 in the UK. DHSC told us that
                             its initial specification for the rapidly manufactured ventilator was created on the
                             assumption that a large number of machines would be required within a very
                             short timeframe. The specification was therefore very basic and focused on
                             key life-saving features. After DHSC published this specification on 14 March,
                             MHRA produced an updated version on 18 March calling for a range of devices
                             from very simple to more sophisticated designs. It updated the specification
                             further during April, reflecting increased clinical understanding of treating
                             COVID-19, issuing the fourth and final version on 10 April. Later iterations of
                             the specification emphasised more sophisticated features such as suction and
                             assisted breathing (paragraph 4.5 and Figure 6).

                             15 Cabinet Office pursued multiple options and shortlisted them using expert
                             advice. Following the Prime Minister’s “call to arms” to UK manufacturers
                             on 16 March and a sift of more than 5,000 initial responses, Cabinet Office
                             convened a ‘technical design authority’ (the TDA) to assess ventilators and inform
                             decisions. The TDA included experts and representatives from the NHS national
                             clinical team, critical care specialists, MHRA and government departments,
                             and drew on data from device-testing experts. The TDA met 12 times between
                             18 March and 21 May. Following its initial meetings at which it rejected some
                             devices, the TDA supported 17 participants and gradually reduced this number
                             as each device proceeded through the regulatory testing process, taking into
                             account the developing picture of demand and government’s targets at the time
                             (paragraphs 4.7 to 4.11 and Figure 6).

                             16 Cabinet Office’s approach was in effect a competition that prioritised
                             speed and maximising the chances of success, before considering cost.
                             The ventilator challenge was not a traditional procurement competition on
                             “most economically advantageous tender” grounds. Instead, the TDA process
                             was a way of continuously assessing multiple options against requirements.
                             Cabinet Office eliminated devices only after it decided they were either: not
                             likely to meet the regulatory standard in time; or, in the end, were not needed.
                             Eventually, Cabinet Office ordered those that met the regulatory standard first.
                             There was no direct competition between participants on cost, although Cabinet
                             Office considered the cost of designs in deciding, for example, the volume and
                             mix of devices (paragraphs 4.9, 4.12 to 4.14, and Figure 6).
Summary 11
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

             17 Cabinet Office accepted higher levels of risk than normal and accepted
             that few designs would meet the regulatory standard in the time available.
             Cabinet Office did not wait to identify which ventilators were most likely to
             work before entering into contracts. Instead, it sought contracts with all the
             participants that remained in the process as if they had been successful,
             issuing conditional letters of intent and agreeing to cover reasonable costs
             where required, until devices were removed via the TDA process. Under these
             agreements, it supported providers to undertake design work, pre-order
             components, develop factory capacity and secure supply chains to ensure the
             ventilators could be built. Cabinet Office committed to covering participants’
             reasonable direct costs and indemnified them against legal actions from
             inadvertently breaching intellectual property rights, competition and procurement
             law, and some aspects of product failure. It estimates it will spend £113 million
             (excluding VAT) on design costs, components and factory capacity for ventilators
             it did not buy because the design was not viable or not needed to meet the
             government’s targets (paragraphs 4.11, 4.15 to 4.19, 4.21 and Figure 7).

             18 Given its overall approach, Cabinet Office took reasonable steps to control
             the programme’s costs where it could. The Cabinet Office sought assurance
             over suppliers’ costs with input from the Ministry of Defence’s Cost Assurance
             and Analysis Service. It also worked with suppliers to cancel unnecessary orders
             early and, where possible, recover costs of components brought in preparation
             for manufacture, selling them back into the wider supply chain. It estimates it
             has recovered about £36.3 million to date in this way. Cabinet Office is also
             working with participants to explore commercial opportunities overseas but
             told us it expects the overall impact of this to be modest in the context of the
             programme’s costs (paragraph 4.20 and Figure 10).
12 Summary
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

                             19 The four ventilator designs Cabinet Office eventually ordered were based
                             on existing designs. It ordered the first devices to successfully meet the MHRA’s
                             requirements. The devices were:

                             •     the Penlon ESO2 from the Ventilator Challenge UK consortium. This is a
                                   modification of an existing anaesthesia machine, designed by Penlon but
                                   built at much greater scale by a group of large UK automotive, aerospace
                                   and other manufacturers;

                             •     variants of the Parapac transport ventilator, an existing design from Smiths
                                   Medical, manufactured at scale by a consortium including Rolls-Royce plc
                                   and GKN Aerospace; and

                             •     two designs from Breas Medical, for which manufacturing was accelerated
                                   and scaled up.

                             Cabinet Office believes that five other new products could have achieved the
                             required regulatory standards given further time. However, during May and
                             June, Cabinet Office decided that it could not justify the cost of further work to
                             support the development of these products as by this time it was clear that the
                             four devices from the three participants above, plus the devices purchased by
                             DHSC, would be sufficient to meet government’s target of 30,000 ventilators
                             (paragraphs 4.12 to 4.14 and Figure 6).

                             The programmes’ costs
                             20 DHSC has spent around £292 million excluding VAT on its ventilation
                             programme. This comprised:

                             •     £244 million for around 11,100 mechanical ventilators, including £221 million
                                   for around 8,100 intensive care unit (ICU) ventilators and £23 million for
                                   around 3,000 transport ventilators;

                             •     £45 million on other oxygen therapy devices; and

                             •     an estimated £3.4 million on programme costs.

                             These costs may understate the actual total because DHSC cannot easily
                             separate out the costs of delivering and distributing the ventilators from overseas
                             because shipping was often combined with other items, such as personal
                             protective equipment (paragraphs 3.14, 3.15 and 3.17, and Figures 5 and 9).
Summary 13
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

             21 Cabinet Office has spent around £277 million excluding VAT on the
             ventilator challenge. This comprises:

             •      £116 million for around 11,700 Penlon ESO2 ventilators;

             •      £26 million for around 1,500 Smiths Parapac ventilators;

             •      £8 million for 2,000 Breas Nippy 4+ and Vivo 65 ventilators;

             •      £113 million on design costs, components and factory capacity for ventilators
                    it did not buy (including around £11 million for an order for 15,000 additional
                    Penlon devices that was later cancelled); and

             •      £14 million on programme costs, which includes around £12 million
                    earmarked for PA Consulting who acted as programme manager, providing
                    specialist knowledge on manufacturing and supply-chain management.

             The final cost of the programme may be lower if Cabinet Office is able to
             recover further costs in the ways described in paragraph 18. It could also be
             higher if any of the participants claim under indemnities Cabinet Office granted
             to protect participants against the risk of product failure and infringement of
             intellectual property rights, although Cabinet Office considers there is a low risk
             of government incurring significant costs in this way (paragraphs 4.19, 4.21 and
             4.23, and Figures 7 and 10).

             22 The cost of the mechanical ventilators acquired varied significantly.
             It is difficult to compare the costs of different machines within or across the
             programmes. All designs have been certified as meeting standards for use in
             the COVID-19 emergency, but they vary widely in their type, functionality and
             clinical utility.

             •      The average total cost of a mechanical ventilator purchased through the
                    ventilator challenge was around £18,300, including programme costs and all
                    the costs of designs that did not proceed to manufacture.

             •      The average cost of mechanical ventilators purchased by DHSC was around
                    £22,300. This included different types of ventilators:

                    •     intensive care ventilators purchased from new suppliers cost an
                          average of around £30,100, compared with an average cost from
                          existing suppliers of around £20,000.

                    •     transport ventilators purchased from new suppliers (including
                          those withdrawn from use), which had an average cost of around
                          £5,300 compared with an average cost from existing suppliers of
                          around £8,800 (paragraphs 3.16 and 4.22, and Figures 5, 9 and 10).
14 Summary
Investigation into how government increased the number of ventilators available to the NHS in response to COVID-19

                             Concluding remarks
                             23 Both Cabinet Office and DHSC started their ventilator programmes on
                             the basis that securing as many mechanical ventilators as possible, as quickly
                             as possible, was necessary to safeguard public health. This urgency was
                             reflected in their approach of: getting the programmes up and running very
                             quickly; protecting their private-sector partners from financial risk; making
                             early commitments to contracts; paying cash upfront for ventilators before
                             they could be inspected; showing a willingness to accept that prices were
                             higher than the normal market rate; deliberately supporting multiple ventilator
                             challenge options; and drawing significantly on technical expertise and capacity
                             from the private sector. In total, the departments spent a total of £569 million
                             across both programmes.

                             24 Ultimately, the anticipated urgent demand for ventilators in mid-April did
                             not materialise. Instead, on 15 April Ministers decided to adopt new targets to
                             provide additional resilience in the system and prepare for a potential second
                             wave. By this point, the departments’ earlier urgency meant that the majority
                             of purchase contracts had been entered into, and the task turned into one of
                             identifying the best mix of devices, ensuring they were delivered, identifying
                             which options were no longer required to meet government’s targets and
                             managing the programmes’ overall cost. While the two departments were not
                             able to meet the initial target of 18,000 mechanical ventilators by the end of April,
                             they made substantial progress towards the later target of 30,000 by the end
                             of June. While the number of ventilators now significantly exceeds demand, this
                             means there is more spare capacity should it ever be needed.

                             25 Inevitably, given the approach the departments took, the overall costs
                             of both programmes are higher than we, or the departments, would expect to
                             see in normal times. However, both departments maintained sufficient record
                             of their programmes’ rationale, the key spending decisions they took and
                             the information they had to base those on. They also put in place effective
                             programme management, controlled costs where they could and recovered
                             some of their committed spending once it became apparent that fewer
                             ventilators were needed than they had originally believed.
You can also read